SOURCE-LINKED INTELLIGENCE

STEVE: Stabilizing Textual Gradient-Based Prompt Optimization via Error-Driven Refinement and Regularized Verification

arXiv · AI, language, vision and robotics · article · Sep 20, 2026 · UTC

Textual-gradient methods automate prompt optimization through natural-language feedback, but their iterative updates can be unstable. We identify two sources of this instability: noisy gradients produced from already-correct examples and over-specialization to hard cases that degrades performance on simpler inputs. We introduce STEVE, a stabilization framework with two coupled mechanisms. Error-Driven Refinement generates gradients only from incorrectly handled examples, concentrating updates on informative failures. Regularized Verification treats every update as provisional and accepts it on

Read original source ↗ Open in workspace

recordType

paper

region

Global